May 2026 Summaries

The EU Data Act is challenging traditional notions of data ownership and control, aiming to give businesses more autonomy over their data by reducing vendor lock-in and encouraging a more open data ecosystem. This legislation directly impacts platforms like SAP, which have historically maintained strict control over data by embedding it within their ecosystems and asserting intellectual property claims over data products. In contrast, the EU Data Act seeks to dismantle these barriers by mandating easier switching between data-processing services, prohibiting unfair contractual clauses that limit data sharing, and clarifying the distinction between a vendor's proprietary software and the data models generated by users. This shift empowers organizations to adopt more flexible data architectures, such as Open Data Infrastructure, which allows for the seamless integration and utilization of data across various platforms without being tethered to a single vendor's ecosystem. Tools like Fivetran facilitate this transition by enabling enterprises to centralize data from systems like SAP into a unified, cloud-based environment, supporting a range of business applications and promoting data accessibility and usability.

Fivetran provides comprehensive solutions for integrating complex healthcare data systems, enabling organizations to create a unified, analytics-ready data foundation. By addressing the challenges posed by fragmented data across platforms like Epic, HL7, EDI, and FHIR, Fivetran simplifies data extraction and integration without the need for fragile, custom-built pipelines. Its pre-built connectors and Hybrid Deployment capabilities allow healthcare organizations to manage clinical, operational, and financial data in compliance with strict regulations, ensuring data is securely processed within their own infrastructure. This results in a robust data integration framework that supports faster decision-making and improved patient outcomes by allowing healthcare teams to focus on insights rather than infrastructure.

Fivetran has revolutionized its customer support operations by integrating AI into its processes, transforming from a human-operated, ticket-by-ticket model to an AI-enhanced system that prioritizes efficiency. This transformation was driven by the challenges of managing complex products and the need for consistent support handovers across global teams. The Fivetran Support AI, a custom-built plugin for Zendesk, leverages the company's extensive data infrastructure, enabling support engineers to access comprehensive, up-to-date knowledge directly within their workflow. This system, powered by data from various sources, allows engineers to ask AI questions, receive draft responses, generate summaries for handovers, and find similar historical tickets. Unlike off-the-shelf AI tools, Fivetran's solution is deeply integrated with its own data ecosystem, ensuring that it provides contextually relevant and accurate support. This AI initiative not only enhances ticket resolution efficiency but also minimizes the manual effort required, positioning humans as overseers rather than processors in the support process.

Open Data Infrastructure (ODI) is increasingly crucial for data leaders as AI-driven workloads require more flexible and scalable data management solutions than traditional architectures can provide. The shift from human-driven to agent-driven queries necessitates an infrastructure that supports continuous, automated operations without being hindered by vendor lock-in or rising data access costs. ODI achieves this by separating storage and compute functions, allowing data to be stored in open file formats within a managed data lake, which can be accessed by various compute engines, including cloud data warehouses, without duplicating data or incurring additional costs. This approach not only enhances flexibility and efficiency but also reduces long-term risk by allowing organizations to adopt new tools and technologies as needed, without being constrained by proprietary systems. Companies like Fivetran facilitate this transition by offering services that automatically ingest data into open formats, enabling organizations to efficiently scale their AI initiatives while maintaining control over their data assets.

As AI agents become the new data consumers, the traditional data stack designed for human analysts is proving inadequate due to its inability to deliver real-time, consistent data necessary for autonomous decision-making. Fivetran and Google Cloud are addressing this challenge by developing an Open Data Infrastructure (ODI), which allows for real-time ingestion and a unified data foundation that is both open and consistent. This architecture supports AI agents by providing them with live, trusted context, enabling them to perform autonomous business actions effectively. The integration with Google's Lakehouse and BigQuery ensures that data is immediately accessible for enhanced processing, such as Retrieval-Augmented Generation, while minimizing latency issues typical of batch processing. Additionally, the use of Knowledge Catalog helps maintain data governance, providing the necessary structure and guardrails for reliable AI operations, thus transitioning from static pipelines to a dynamic, always-on data environment that aligns with open standards and automated ingestion.

Fivetran has effectively utilized AI internally by building tools that enhance productivity through a centralized data infrastructure, which is key to reaping the benefits of AI. By ensuring their data is centralized, standardized, and well-structured, Fivetran has been able to integrate AI into their workflows to improve customer request management, service visibility, and product change notifications. This centralized approach prevents data silos and rogue analytics, allowing for consistent metrics and efficient data processing, which supports scalable AI applications. The company emphasizes that the success of AI initiatives is less about having the most advanced AI models and more about having a robust data foundation that ensures data is readily available and in context. This strategy has led to significant time savings and improved decision-making, reinforcing the importance of a strong data infrastructure as companies increasingly adopt AI technologies.

In an effort to optimize product operations, the Chief Product Officer at Fivetran explored the use of AI agents and conversational analytics to enhance productivity and resource allocation within the company. By integrating Jira data into BigQuery using Fivetran and leveraging generative and agentic AI through tools like Claude, a more streamlined and efficient approach to data analysis was achieved. This method allowed the team to bypass traditional analytics processes that involved manual dashboard creation and SQL usage, instead enabling queries in natural language that translated into actionable insights. The initiative revealed critical insights into team performance and dependency issues, significantly improving project prioritization and planning. The experience highlighted the importance of maintaining a centralized, interoperable data infrastructure for analytics and demonstrated how AI can democratize data access by simplifying the querying process, thus enabling informed decision-making across teams.

As organizations transition from traditional analytics to leveraging artificial intelligence (AI), the limitations of the modern data stack, which is characterized by tightly coupled, warehouse-centric architectures, become evident. This architecture, initially effective for consolidating and querying data, now struggles with the demands of AI, leading to increased costs and latency issues. The proposed solution is to shift towards an Open Data Infrastructure (ODI), which decouples storage from compute by utilizing open formats like Iceberg and Delta, allowing various platforms such as Snowflake and Databricks to access a unified data source. This approach minimizes duplicated pipelines, reduces costs, and maintains data consistency across different teams and tools. While ODI introduces some operational complexities, it offers greater flexibility and scalability for evolving workloads, particularly in AI applications, by enabling data portability and reusability without vendor lock-in. Despite past failures with data lakes, advancements in open table formats and managed data services now support more successful implementations. The shift is not about abandoning existing platforms but repositioning data to optimize its accessibility and utility across multiple engines, thereby future-proofing data strategies in a rapidly evolving technological landscape.

Modern data teams require flexible and open data infrastructure to accommodate the diverse needs of various roles such as analysts, data engineers, ML engineers, and data scientists, who each utilize different tools and workflows. Traditional centralized data systems fail to meet these needs, often resulting in data duplication, rising costs, and outdated data due to the constraints of a single-platform approach. Open Data Infrastructure (ODI) addresses these challenges by decoupling storage from compute, utilizing open standards like Apache Iceberg and Delta Lake, allowing multiple engines to operate on a single data set without duplication. Fivetran exemplifies ODI by providing a Managed Data Lake Service that ensures reliable, automated data ingestion in open formats, which prevents lock-in and reduces maintenance overhead, facilitating seamless tool integration and efficient data management. This approach supports a dynamic, scalable data ecosystem where each team can use their preferred tools without re-architecting the data stack, enhancing flexibility and experimentation capabilities.

To gain a competitive edge in AI readiness, organizations should focus on centralizing their data infrastructure to ensure it is complete, current, and accessible, as the real advantage lies not in the AI models themselves, which are widely available, but in the data quality and infrastructure. The text outlines a strategic approach to preparing for AI, emphasizing the urgency of quantifying data bottlenecks, centralizing data using modern tools like Fivetran, and building robust pipelines with change data capture (CDC) capabilities to support agentic AI. It highlights the importance of storing data in open, portable formats like Apache Iceberg to avoid costly migrations and ensure flexibility across future technological shifts. Additionally, organizations are encouraged to think beyond their current data sources, identifying new data categories that could significantly enhance their AI capabilities. The text underscores that the timing of these strategic moves is critical, as early adopters are likely to establish a substantial lead in AI-driven competitive advantages.

Fivetran's "Agentic AI readiness index 2026" highlights that while enterprises are heavily investing in agentic AI, a significant 85% lack the necessary data foundation to support it at scale, leading to potential widespread operational failures. Despite 41% of companies already running agentic AI in production, issues like data quality, lineage, sovereignty, and compliance are major obstacles, as identified by 42% and 39% of respondents respectively. The research indicates that only 15% of organizations are fully prepared, and these organizations report high confidence in their AI's return on investment. The key to successful AI implementation lies in having an Open Data Infrastructure, which ensures consistent, governed data movement, interoperability, and clear governance, allowing AI systems to function reliably without human oversight. This infrastructure transition transforms AI from a mere productivity tool to an integral part of the operational model, differentiating successful companies from those struggling to realize AI benefits.

SQLGlot is a Python SQL parser, transpiler, and optimizer designed to support 34 SQL dialects with zero dependencies, making it popular for projects like SQLMesh and Apache Superset. Its primary advantage is its ease of use, attributed to its pure Python nature, though this typically comes at the expense of speed. To enhance performance without abandoning its Python roots, SQLGlot adopted mypyc, a tool that transforms type-annotated Python code into C extension modules, which led to significant speed improvements in processing SQL queries. Despite challenges, including numerous bugs within mypyc and difficulties in handling SQLGlot's large codebase, the integration was successful, yielding speedups such as a 5x increase in parsing speed. However, SQLGlot continues to offer a pure Python version to accommodate users with specific needs, such as the inability to build C extensions. The transition to mypyc not only improved performance but also uncovered bugs in the codebase, encouraging a more robust development process. Additionally, SQLGlot contributed to mypyc's development by fixing internal compiler issues, thus benefiting the broader Python community.